The overall goal of this application is to understand relationships among hormonal, genetic and behavioral risk factors for symptomatic benign prostatic hyperplasia (BPH). BPH symptoms affect well over half of American men over age 50, with substantial effects on quality of life and medical care cost of over 4 billion dollars annually. This study will assess whether the risk of symptomatic BPH is associated with (1) serum concentrations of steroid hormones and insulin-like growth factors known to affect prostate growth; (2) polymorphisms in genes that affect hormone and growth factor metabolism or activity; and (3) diet and other lifestyle factors that affect steroid hormones and insulin-like growth factor concentration or activity. Data are from the Prostate Cancer Prevention Trial (PCPT), a double blinded, placebo-controlled, randomized trial of the drug finasteride (Proscar) for the primary prevention of prostate cancer. Analyses will be restricted to the approximate 5,000 men in the placebo control group with no medical history or symptoms of BPH at baseline. The primary endpoint, incidence of symptomatic BPH, is based on careful and standardized assessments of lower urinary tract symptoms and treatment for prostate-related disease collected at baseline and annually for 7 years. Steroid hormones, insulin-like growth factor 1 and insulin-like growth factor binding protein 3 will be measured from plasma banked at baseline, and genetic characteristics will be assessed using banked lymphocytes. Diet and other behavioral risk factors were assessed by both interviewer- and self-administered questionnaires. A nested case-control design, with 700 cases and 700 controls, will be used to address hypotheses related to serum measures and genetic polymorphisms. A cohort study using all 5,000 men will be used to examine hypotheses related to diet and other behavioral risk factors. The nested case-control studies are powered to detect odds ratios of 1.5 and the cohort studies are powered to detect hazard ratios of 1.35, comparing highest to lowest quartiles of exposures, with type I error = 0.05 and type II error = 0.20. This study is highly cost-effective, as all data and samples have been collected, and funds are requested for laboratory analyses of serum and genetic polymorphisms, data preparation, statistical analyses and manuscript preparation. Results from this study will enhance our understanding of the etiology of symptomatic BPH and could potentially be used to develop new strategies for prevention and control of this very common disease.